Similarity Predictions: 2nd Round (Sunday)

I'll start with the prediction chart for all of Sunday's games, then get to the previews for the Big 12 games, which are also cross posted over at Upon Further Review.  If you are unfamiliar with my similarity predictions, read about the methods here.  Also, sorry about the ugly tables - I tried a new method of posting them, and it didn't work so well.  You live, you learn.

SIMILARITY vs. POMEROY
GAME			POMEROY	SIMILARITY
Syracuse	-	Gonzaga	Syr +8.5 (78%)	Syr +11 (84%)
Ohio St	-	Georgia Tech	OSU +4.5 (68%)	OSU +7.5 (77%)
Maryland	-	Michigan St	MD +3.5 (64%)	MD +9 (81%)
West Virginia	-	Missouri	WVU +2.5 (60%)	MU +3 (62%)
Wisconsin	-	Cornell	Wisc +8.5 (82%)	Corn +2.5 (61%)
Pittsburgh	-	Xavier	Xav +1.5 (56%)	Xav +2 (57%)
Purdue	-	Texas A&M	Pur +0.5 (51%)	Pur +4.5 (69%)
Purdue w/o Hummel	-	Texas A&M	Pur +0.5 (51%)	A&M +3.5 (79%)
Duke	-	California	Duke +8 (78%)	Duke +9.5 (82%)

Despite my predictions of glory, Oklahoma State fell to Georgia Tech on Friday*, leaving the Big 12 with only two games on Sunday.  As before, I’ll go ahead and assume readers of UFR are familiar with the Big 12 participant, so these previews will focus on West Virginia and Purdue.

*This is a good time to point out the role of luck in determining whether a prediction looks good.  Georgia Tech made 24 of 25 free throws on Friday, despite a season average of only 64%.  Free throw shooting should be the single thing that match up issues do not affect, as nobody’s defending a FT.  If GT had missed their shots at the usual rate, they’d have ended up with 8 less points, OSU would be playing on Sunday, and I’d look a tad smarter.

MISSOURI vs. WEST VIRGINIA

West Virginia got off to a slow start on Friday, falling behind 10-0 and missing their first 11 shots, but from that point on they outscored Morgan State by 37.  This was more a product of defense than offense, as the Mountaineers only shot 27% from deep and 42% overall, but held the Bears to even lower marks (16% and 29%).  That’s a departure from their norm, as West Virginia’s offense is better than their defense, at least according to adjusted efficiency.  Let’s take a look at that offense:

WEST VIRGINIA OFFENSE
Key Traits:			Off Eff	OReb%	TO%	2P%	3P%
	2010	West Virginia	116.8	41.9	18.1	48.7	33.5
Std Dev from Ave			1.7	2.4	1.3	0.2	-0.3
SIM	Similar Teams		Off Eff	OReb%	TO%	2P%	3P%
92	2010	Providence	115.4	41.3	17.3	49.2	33.0
91	2010	Kansas St.	116.4	40.9	20.4	49.4	35.7
89	2010	Kentucky	116.1	40.1	20.3	53.9	34.7
89	2010	Louisville	113.9	38.9	19.8	51.9	33.7
88	2010	Michigan St.	111.7	40.2	21.1	51.7	33.5

West Virginia plays strong – strong on the offensive glass (best OReb% in the country), and strong with the ball (59^th-lowest TO%).  But they have to rely on their strength, as they don’t shoot well from anywhere.  Part of their low 2P% can probably be chalked up to the lack of a true point guard (their assists leader is Da’Sean Butler); the lack of leadership and creativity means they likely take more difficult shots than some teams.

WEST VIRGINIA DEFENSE
Key Traits:			Def Eff	OReb%	eFG%	Steal%	2P%
	2010	West Virginia	89.4	30.6	47.1	8.6	46.2
Std Dev from Ave			1.7	1.0	0.8	-0.8	0.6
SIM	Similar Teams		Def Eff	OReb%	eFG%	Steal%	2P%
95	2010	Texas A&M	89.2	30.2	46.8	9.9	45.0
92	2010	Pittsburgh	90.5	30.7	44.3	8.2	43.5
90	2010	Ohio St.	89.7	28.7	47.0	10.9	44.5
90	2010	Texas El Paso	89.1	30.1	44.6	12.3	43.4
90	2010	Texas	90.4	30.6	44.9	10.6	43.6

Here again, WVU’s emphasis of toughness over finesse is evident - they’re better than average in rebounding, they foul a lot, and they hold opponents to a subpar eFG% (via physical annoyance?), but they don’t pick many pockets.  Missouri already played a carbon copy of West Virginia’s defense earlier this year, in Texas A&M – they lost the game, but their offense actually performed well, considering the quality of the defense they were up against. 

Now let’s take a look at each team’s performance against similar opponents.  I’ve changed the presentation format once again – this is still a work in progress.  My guess is you readers thought there were too many charts before, so I’ve eliminated those that show single game performance against the top ten comps.  Now, I just summarize that data in the top half of the table below, and the bottom half is the game prediction.  This way, you can see at a glance what the prediction is, and which aspects of team performance are most influencing it, but you don’t have to wade through a rainbow of numbers:

MISSOURI vs. WEST VIRGINIA
Adjusted Efficiency
Team/Unit	Full Season	Vs. Comps
Missouri Off	111.4	116.4
Missouri Def	88.4	91.1
West Virginia Off	116.8	116.3
West Virginia Def	89.4	94.9
Game Prediction
System:	Pomeroy	Similarity
Winner	West Virginia	Missouri
Margin	+2.4	+2.9
Probability	60%	62%

OK, I lied – there’s still a rainbow, just a much smaller one.  In the top section, the color coding indicates whether a team did better (green) or worse (red) than their season average when facing the top ten most similar opponents. The bottom half is the same as before, and hopefully self explanatory. Onward…

MU’s offense has in general done well against physical defenses – they were better than their season average against 8 of the 10 relevant opponents.  In most of these cases, no one category stood out as being dominated by MU – they simply did a good job of maintaining their usual level of play, despite facing a good defense.

The other large difference here is in defensive efficiency, with West Virginia coming in at 4 points worse than their season average.  The main culprits are a low number of turnovers and a high number of free throws, so it looks like Missouri should be able to hold onto the ball and get to the line.  They’ll need to do as well as they did against Clemson (9 less TO and 15 more FTA) if they want to knock off West Virginia.

The two predictions here are complete opposites, and my system apparently agrees with Stewart Mandel in thinking that the Tigers may be a bad matchup for the Mountaineers.  Let’s hope it’s right,  both for Missouri’s sake and for mine, as I imagine I’ll be judged largely on how a high profile upset pick such as this does.

TEXAS A&M vs. PURDUE

Purdue’s obviously not the same team without Robbie Hummel, so I’m going to try something new here.  I’m going to run two tables each for Purdue’s offense and defense – one using only pre-Hummel stats, and one using post-Hummel stats. That way we can compare them and find out what the loss of Hummel has done to the Boilermakers.

PURDUE OFFENSE with Hummel
Key Traits:			Off Eff	TO%	eFG%	OReb%	FTA/FGA
	2010	Purdue	115.6	16.6	51.1	32.1	39.7
Std Dev from Ave			1.6	1.9	0.6	-0.3	0.6
SIM	Similar Teams		Off Eff	TO%	eFG%	OReb%	FTA/FGA
95	2010	New Mexico	114.3	16.4	51.0	35.7	45.6
94	2010	Memphis	114.1	17.0	53.3	32.8	44.8
94	2010	Marquette	114.3	15.8	52.5	30.9	35.1
93	2010	Maryland	119.1	16.7	51.9	35.4	33.9
93	2010	Northwestern	113.0	17.3	52.2	30.3	35.7

PURDUE OFFENSE without Hummel
Key Traits:			OReb%	eFG%	TO%	FTA/FGA	Off Eff
	2010	Purdue	22.5	43.6	16.4	47.6	102.6
Std Dev from Ave			-2.9	-1.8	2.0	2.2	0.1
SIM	Similar Teams		OReb%	eFG%	TO%	FTA/FGA	Off Eff
84	2010	MD Baltimore County	24.2	45.8	17.6	32.7	93.0
82	2010	Texas Pan American	23.5	44.8	20.7	36.8	86.3
81	2010	Saint Joseph's	26.9	46.6	18.5	38.7	100.2
80	2010	Manhattan	30.7	44.2	17.2	36.1	97.3
79	2010	Illinois Chicago	28.0	43.7	19.3	33.0	94.6

YIKES.  His injury has absolutely destroyed their offense.  They were already a below average rebounding team, and now they’re absolutely getting killed on the boards.  Hummel was also their best high-volume shooter, so their eFG% has plummeted as expected.  The one positive is that they are getting fouled more, but that’s the least important number on the chart.  The change is a bit less damaging on defense:

PURDUE DEFENSE with Hummel
Key Traits:			Def Eff	eFG%	OReb%	TO%	FTA/FGA
	2010	Purdue	87.4	45.4	29.7	23.6	38.6
Std Dev from Ave			2.0	1.5	1.3	1.2	-0.3
SIM	Similar Teams		Def Eff	eFG%	OReb%	TO%	FTA/FGA
96	2010	Tennessee	87.3	44.9	31.5	23.7	40.9
96	2010	Old Dominion	88.5	45.3	29.4	22.8	30.5
94	2010	Texas El Paso	88.3	44.2	30.2	22.4	42.9
93	2010	Nevada Las Vegas	90.7	45.5	30.4	22.9	47.1
92	2010	Dayton	88.8	45.3	28.0	21.4	43.4

PURDUE DEFENSE without Hummel
Key Traits:			Def Eff	eFG%	TO%	OReb%	FTA/FGA
	2010	Purdue	86.8	44.5	26.2	35.9	32.8
Std Dev from Ave			2.1	1.8	2.3	-0.9	0.7
SIM	Similar Teams		Def Eff	eFG%	TO%	OReb%	FTA/FGA
91	2010	Missouri	88.4	45.9	27.3	37.7	38.1
89	2010	Purdue	86.2	45.0	23.8	31.0	36.9
89	2010	Clemson	88.8	46.4	24.7	33.6	35.4
89	2010	Tennessee	87.3	44.8	23.6	31.6	40.8
88	2010	Virginia Tech	88.1	44.3	22.4	33.0	35.2

They’ve actually been a slightly better defensive team without Hummel, but only by a point, not anywhere near enough to offset the offensive drop.  They’re allowing more offensive rebounds, which makes sense, as Hummel had their highest defensive rebound percentage.  But they’re fouling less, and – here is the big key – causing a lot more turnovers.  They’ve changed their identity, but not lost any quality.

As for the prediction, I don’t really know what to do with this one.  There’s not enough data to use my system on the sans-Hummel version of Purdue, so I’ll show you numbers that are derived from full season stats, and slap a huge warning label on it.

TEXAS A&M vs. PURDUE
Adjusted Efficiency
Team/Unit	Full Season	Vs. Comps
Texas A&M Off	112.6	109.4
Texas A&M Def	89.2	89.5
Purdue Off	109.3	113.2
Purdue Def	86.2	86.5
Game Prediction
System:	Pomeroy	Similarity
Winner	Purdue	Purdue
Margin	+0.3	+4.4
Probability	51%	69%
DO NOT TRUST THIS!

What, you thought I was kidding?  Given that Pomeroy’s prediction is a toss up, despite giving Purdue credit for fantastic pre-Hummel-injury performance, I feel comfortable saying that Texas A&M should be the clear favorite against the current Purdue squad.

Whether you'd like to shower me with accolades or draft articles of impeachment, feel free to email me!